The treatment of tuberculosis in persons co-infected with HIV using standard rifampin-based regimens is generally successful. However, rifampin activates the hepatic cytochrome p450 enzyme system, resulting in a 40-90% decrease in serum concentrations of the HIV-1 protease inhibitors, a key component of highly active antiretroviral therapy. Rifabutin ishighly active against M. tuberculosis, yet is a less-potent inducer of the cytochrome p450 enzyme system than rifampin. The CDC Tuberculosis Trials Consortium study, "Treatment of HIV-related tuberculosis using a rifabutin-based regimen" (USPHS Study 23--JCCI RPN 99-01-26-01), is evaluating the efficacy of an intermittent TB treatment regimen that includes rifabutin. Although the pharmacokinetics of isoniazid and rifabutin in immunocompetent individuals are sufficient to result in effective therapy without drug toxicity, there are concerns that malabsorption and diarrhea in HIV-infected persons will affect the pharmacokinetics of both drugs. The purpose of this study is to determine the proportion of patients with HIV-related tuberculosis who have abnormal pharmacokinetics of isoniazid and rifabutin. This study will also determine risk factors for abnormal pharmacokinetics of isoniazid and rifabutin, and evaluate the correlation between the pharmacokinetic parameters of isoniazid and rifabutin and the therapeutic efficacy and toxicity attributed to these medications.